We report a single-cell bisulfite sequencing method (scBS-Seq) capable of accurately measuring DNA methylation at up to 48.4% of CpGs. We observed that ESCs grown in serum or 2i both display epigenetic heterogeneity, with “2i-like” cells present in serum cultures. In silico integration of 12 individual mouse oocyte datasets largely recapitulates the whole DNA methylome, making scBS-Seq a versatile tool to explore DNA methylation in rare cells and heterogeneous populations.

Technological advances in genomics and imaging have led to an explosion of molecular and cellular profiling data from large numbers of samples. This rapid increase in biological data dimension and acquisition rate is challenging conventional analysis strategies. Modern machine learning methods, such as deep learning, promise to leverage very large data sets for finding hidden structure within them, and for making accurate predictions. In this review, we discuss applications of this new breed of analysis approaches in regulatory genomics and cellular imaging. We provide background of what deep learning is, and the settings in which it can be successfully applied to derive biological insights. In addition to presenting specific applications and providing tips for practical use, we also highlight possible pitfalls and limitations to guide computational biologists when and how to make the most use of this new technology.

We report scM&T-seq, a method for parallel single-cell genome-wide methylome and transcriptome sequencing, allowing discovery of associations between transcriptional and epigenetic variation. Profiling of 61 mouse embryonic stem cells confirmed known links between DNA methylation and transcription. Notably, the method reveals novel associations between heterogeneously methylated distal regulatory elements and transcription of key pluripotency genes.

Recent technological advances have enabled DNA methylation to be assayed at single-cell resolution. However, current protocols are limited by incomplete CpG coverage and hence methods to predict missing methylation states are critical to enable genome-wide analyses. We report DeepCpG, a computational approach based on deep neural networks to predict methylation states in single cells. We evaluate DeepCpG on single-cell methylation data from five cell types generated using alternative sequencing protocols. DeepCpG yields substantially more accurate predictions than previous methods. Additionally, we show that the model parameters can be interpreted, thereby providing insights into how sequence composition affects methylation variability.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-017-1189-z) contains supplementary material, which is available to authorized users.